CN101288045B - Enhanced host interface - Google Patents

Abstract

A memory system that is compatible with hosts using different protocols includes protocol adapters for the different protocols. Protocol adapters allow a common backend system to be used for data that is provided in different formats. The protocol adapter generates responses to a host and generates commands for a backend as appropriate.

Description

Enhanced host interface

Technical field

Do not have

Background technology

The application's case relates to the for example operation of the re-programmable Nonvolatile memory system of semiconductor flash memory, and more particularly, relates to the management of the interface between host apparatus and the storer.All patents that this paper quotes, patent application case, article and other publication, document and things (comprising all application cases that preceding text are quoted under " cross reference of related application ") are incorporated herein in this mode of quoting in full from all purposes.

In the commercial flash memory system of generation morning, the rectangular memory cell array is divided into a large amount of cell group, the data volume of each storage standards disk drive sector, i.e. 512 bytes.Also comprise the for example additional data volume of 16 bytes in each group usually, with storage error correction sign indicating number (ECC) and possible and user data and/or other relevant overhead data of group of memory cells of storaging user data wherein.Memory cell in each this group is the minimum number destination memory location that can wipe jointly.That is to say that erase unit is a large amount of memory cells of data sector of storage and any overhead data that is comprised effectively.The case description of the accumulator system of this type is in the 5th, 602, and No. 987 and the 6th, 426, in No. 893 United States Patent (USP)s.Need before it being programmed again, wipe the characteristic that memory cell is a flash memory with data.

The most common form with memory cards or flash drive provides flash memory system, and said memory cards or flash drive are connected with multiple main frame such as for example personal computer, camera etc. removedly, but also can be embedded in these host computer systems.When storer writes data, main frame usually the sector in the continuously and virtually address space of accumulator system, troop or other data unit is assigned unique logical address.Be similar to disc operating system (DOS) (DOS), the address of main frame in the logical address space of accumulator system writes data and from its reading of data.Controller in the accumulator system will translate to the physical address in the actual storage memory of data array from the logical address that main frame receives, and then follow the tracks of these address translations.The data storage capacity of accumulator system is the same big with addressable data volume on the whole logical address space that defines for accumulator system at least.

In later several generations flash memory system, the size of erase unit is increased to the block of enough memory cells to store a plurality of data sector.Although the host computer system that is connected with the accumulator system for example less least unit of sector is programmed and reading of data, a large amount of sector storage are in the single erase unit of flash memory.Commonly some data sector in the block become out-of-date when main frame renewal or replacement logic data sector.Owing to wipe whole block before any data that must in can be to block, store override, therefore new or data updated is stored in another block of having wiped and having had the residual capacity that is used for data usually.This process has emptied and has had the original block that occupies the outdated data in valuable space in the storer.If but wherein remained with any valid data, said block could be wiped so.

Therefore; In order to utilize the memory capacity of storer better; Usually merge or collect live part block data amount, this accomplishes in the following manner: copy data in the block of wiping, make data that these duplicate from block then can be through wiping and its whole memory capacity can re-use.Also expect data are duplicated so that it divides into groups by the inferior ordered pair of the logical address of data sector in the block, because this speed that increases reading of data and the data that read are sent to main frame.If the generation that this data are duplicated is too frequent, the operating performance of accumulator system possibly demoted so.Memory capacity at storer does not almost have under the situation (typical case) of difference through the addressable data volume of the logical address space of system with main frame, and this especially influences the operation of accumulator system.In the case, before can carrying out the main frame program command, possibly carry out data merging or collection.Therefore the programming time increases.

In the constant generations accumulator system, the size of block increases gradually, so that increase the number of storable data bit in the given semiconductor area.Store 256 data sectors and more the block of multidata sector become common.In addition, usually with different arrays or subarray two, four or more multi-tiling be linked in logic and become first block together so that the concurrency degree when increasing data programing and reading.What occur with this high capacity operating unit is the challenge that operates on it effectively.

Along with new innovation allows bigger memory span and speed, need provide usually and use this type of innovation but still the product compatible with the product that does not use this type of innovation.This means new product technical renovation capable of using but still back compatible, make it to use with the product that uses old technology.This back compatible property is even more important for the portable product that can in having the various configurations of multiple different technologies, use.An instance of this portable product is removable flash memory cards.

Summary of the invention

A kind of accumulator system comprises with the interface layer of main-machine communication and with the rear end of data storage in memory array.Between said interface layer and said rear end, translation layer will be the intelligible form in rear end by data and the command conversion that said interface layer receives according to the different agreement that main frame uses.Therefore, translation layer allows rear end commonly used to use with the multiple main frame that uses various protocols.This is particularly useful for the accumulator system in the removable memory cards.Translation layer contains one or more protocol adaptors.Protocol adaptor receives the communication (order and data) from main frame according to the employed agreement of main frame, and as responding to changing said data and order in the rear end.Protocol adaptor also can produce signal to main frame, and wherein these signals are parts of host protocol.

A kind of accumulator system comprises the object protocol adaptor, its will become with command conversion according to the data that the object agreement is sent and nonvolatile memory in the form based on the memory storage compatibility of file.In particular, the object protocol adaptor received the metadata about object before receiving object.The size of object is included in the metadata.The object protocol adaptor determines when and receives whole object through reception data volume and the indicated size of metadata information are compared.When the object protocol adaptor confirmed to have received whole object, it produced to the response of main frame and to the rear end of accumulator system and produces end of file indicator, makes file closed by the rear end.This allows rear end schedule file carrying out refuse collection, and thereby allows more effectively to store and file control data.

A kind of accumulator system comprises the LBA protocol adaptor, its will become with command conversion according to the data of LBA agreement with nonvolatile memory in based on the memory storage of file compatible data and order.The data map of the logical address of in an example, having of being received being assigned from the logical address space that defines for accumulator system by main frame is to logical file.Logical file is then handled with the mode identical with other file by the rear end.Logical file occupies whole first block usually, make its not can with other data sharing unit block.Yet identical block can be used for logical file and be used for other file At All Other Times a time, makes memory array be listed in not hard division the between the dissimilar files.

A kind of accumulator system include file protocol adaptor, its will become with command conversion according to the data of file protocol with nonvolatile memory in based on the memory storage of file compatible data and order.Under the situation of the rear end use agreement identical (for example, all using the immediate data file protocol), possibly needn't translate with main frame.Yet use at main frame under the situation of different files agreement, the file protocol adapter is made suitable translating.

In some cases, accumulator system maybe with an above main-machine communication that uses an above protocol adaptor.Under said situation, translation layer can select a protocol adaptor to come and back end communications at every turn.In some cases, translation layer can be arbitrated between different main frames through alternate selection different protocol adapter, so that the interleaving access to the rear end to be provided.

Interface layer comprises the logic interfacing with a plurality of compatible mainframes.In some cases, possibly exist independent physical interface be used for host apparatus on the corresponding interface be connected.Yet this is dispensable, and in some cases, the single physical interface of USB connector for example is provided and is used by all logic interfacings.Interface layer and translation layer function can be carried out or can be carried out by the firmware on the controller by special circuit.This can be the Memory Controller of diode-capacitor storage array.Memory array can be the nand memory array and can be formed on one or more semi-conductor chips.Accumulator system can be included at different time and be connected in the removable card of different main frames.

Back-end system can be managed data as file, said file is in some cases corresponding to Hosts file (but in some cases, not with Hosts file corresponding one by one).Based on an instance of the back-end system of file is the immediate data file rear end of describing in the 11/060th, 174,11/060,248 and 11/060, No. 249 patent application and the 60/705th, No. 388 temporary patent application case.

Description of drawings

Main frame and the Nonvolatile memory system that is connected that Fig. 1 schematic illustration is implemented at present;

Fig. 2 is the block diagram as the exemplary flash memory system of the nonvolatile memory of Fig. 1;

Fig. 3 is the representative circuit diagram of the memory cell array that can in the system of Fig. 2, use;

The exemplary physical memory organization of the system of Fig. 4 key diagram 2;

The expander graphs of the part of the physical storage of Fig. 5 exploded view 4;

Fig. 6 show Figure 4 and 5 physical storage a part further expand figure;

Fig. 7 explains the common prior art logical address interface between main frame and the re-programmable accumulator system;

Fig. 8 explains the common prior art logical address interface between main frame and the re-programmable accumulator system with the mode different with Fig. 7;

Fig. 9 explanation is according to the direct file memory interface between main frame of the present invention and the re-programmable accumulator system;

Figure 10 explains according to the direct file memory interface between main frame of the present invention and the re-programmable accumulator system with the mode different with Fig. 9;

Figure 11 shows the scheme of the Hosts file be used to store as the sector of the logical address with shared logical address space that the storer of doing for oneself defines and receive from main frame; Said sector map then is stored in the memory array to logical file and logical file, wherein logical file of each first block;

Figure 12 is used for the logic address data is stored in the scheme in the logical file of memory array with the mode explanation that is different from Figure 11.

Figure 13 show have with based on the file interface of the back end communications of file and the accumulator system of LBA interface, wherein the LBA protocol adaptor is inserted between LBA interface and the rear end based on file.

Figure 14 A shows MTP " transmission object information " affairs.

Figure 14 B shows MTP " transmission object " affairs.

Figure 15 shows to have object interface and based on the accumulator system of the rear end of file, wherein the object protocol adaptor is inserted between object interface and the rear end based on file.

Figure 16 shows to have the accumulator system of file interface, object interface and LBA interface, its have file protocol adapter, object protocol adaptor and LBA protocol adaptor so that interface with based on communicating by letter between the rear end of file.

Figure 17 shows the alternative view of the accumulator system of Figure 16, wherein file interface, object interface and LBA interface is regarded as the part of interface layer, and file protocol adapter, object protocol adaptor and LBA protocol adaptor are regarded as the part of translation layer.

Embodiment

The big volume description of flash memory

With respect to Fig. 1 to Fig. 8 describe current flash memory system and with the typical operation of host apparatus.But the various aspects of embodiment of the present invention in this system.The host computer system 1 of Fig. 1 with data storage in flash memory 2 and from retrieve data wherein.Although flash memory can be embedded in the main frame, storer 2 is illustrated as the card than common form, its mating part 3 and 4 through machinery and electrical cnnector is connected to main frame removedly.Have many different commercially available flash memory cards at present, instance is compact flash memory (CF), multimedia card (MMC), secure digital (SD), small-sized SD, memory stick, smart media and TransFlash card.Although each in these cards all has machinery and/or electric interfaces according to the uniqueness of its standard convention, the flash memory that comprises in each is very similar.These cards can be buied from the assignee SanDisk company of the application's case.It is the flash drive line of trade mark that SanDisk also provides with Cruzer, and it is the hand-held accumulator system of little encapsulation, and said system has and is used for the USB plug that is connected with main frame through USB (USB) socket that inserts main frame.In these memory cards and the flash drive each all comprises the controller that connects and control the operation of flash memory in it with main frame Jie.

Use the host computer system of these memory cards and flash drive to have many and of all kinds.They comprise personal computer (PC), on knee and other portable computer, cellular phone, PDA(Personal Digital Assistant), Digital Still Camera, digital movie camera and portable audio player.Main frame comprises the memory cards that is used for one or more types or the built-in socket of flash drive usually, but some of them need to insert the adapter of memory cards.

With regard to the scope that storer 2 relates to, the host computer system 1 of Fig. 1 can be regarded as has two major parts, by constituting of circuit and software.They are application program part 5 and the driver part 6 that is situated between and connects with storer 2.In personal computer for example, application program part 5 can comprise the processor of operation word processing, figure, control or other popular application software.Mainly be exclusively used in the host computer system of carrying out the simple function set at camera, cellular phone or other, application program part 5 comprise the operation camera with shooting with storage picture, operate the software of cellular phone to make a phone call and to answer the call or the like.

The accumulator system 2 of Fig. 1 comprises flash memory 7 and circuit 8, and circuit 8 and be situated between with the main frame that is used for back and forth Data transmission with card connection connects, and goes back control store 7.Controller 8 is changed between the physical address of the logical address of the data that main frame 1 uses and storer 7 at data programing and during reading usually.

Referring to Fig. 2, its description can be used as the circuit of typical flash memory system of the nonvolatile memory 2 of Fig. 1.System controller is implemented on the single integrated circuit chip 11 usually, and it is connected in parallel through system bus 13 and one or more integrated circuit memory chips, shows single sort memory chip 15 among Fig. 2.Illustrated specific bus 13 comprises independent one group of conductor 17 of being used to carry data, be used for one group of conductor 19 of storage address and be used to control and one group of conductor 21 of status signal.Perhaps, single one group of conductor can be between these three functions the time share.In addition; Can adopt other configuration of system bus; For example the title of application on August 9th, 2004 is the ring bus of describing in the 10/915th, No. 039 patent application of " Ring Bus Structure andIt ' s Use in Flash Memory Systems ".

Typical control device chip 11 has the internal bus 23 of himself, and internal bus 23 is situated between with system bus 13 through interface circuit 25 and connects.The major function part that is typically connected to bus is processor 27 (for example microprocessor or microcontroller), contains the ROM (read-only memory) (ROM) 29 of the code that is useful on initialization (" guiding ") system, is mainly used in the random-access memory (ram) 31 that the data that transmit between storer and the main frame are cushioned, and to being used to transmit the circuit 33 that calculates and check through the error-correcting code (ECC) of the data of the controller between storer and the main frame.Controller bus 23 is situated between with host computer system through circuit 35 and connects, and is included under the situation in the memory cards in the system of Fig. 2, and institute gives an account of and connects is that the external connector 37 as the part of connector 4 through blocking is accomplished.Clock 39 is connected with each other assembly of controller 11 and by its use.

Memory chip 15 and any chip that other is connected with system bus 13 contain the memory cell array that is organized into a plurality of subarrays or plane usually; Two this planes 41 and 43 are described for brevity; But the more planes of alternative use, for example four or eight this planes.Perhaps, can be not the memory cell array of chip 15 be divided into the plane.Yet when so dividing, each plane has the arrange control circuit that can operate independently of one another 45 and 47 of himself. Circuit 45 and 47 receives its address of memory cell array separately from the address portions of system bus 13 19, and it is decoded specific one or more in the bit line 49 and 51 are separately carried out addressing.Addressed word line 53 is come through line control circuit 55 in address in response on address bus 19, receiving.Source voltage control circuit 57 also is connected with separately plane with 59, as p trap voltage control circuit 61 and 63.If memory chip 15 has the single memory cell array, and if have two or more this chips in the system, can be similar to the array that interior plane of above-mentioned many planar chip or subarray are operated each chip so.

Data transmit into and send out plane 41 and 43 through the data input/output circuit separately 65 and 67 that is connected with the data division 17 of system bus 13.Circuit 65 and 67 is used for through be connected to the line 69 and 71 on plane via arrange control circuit 45 and 47 separately, programs data in the memory cell and from its memory cell reading of data on plane separately.

Although the operation of controller 11 control store chips 15 comes programming data, reading of data, wipes and look after various housekeeping incidents, each memory chip also contains certain control circuit, and its order of carrying out self-controller 11 is to carry out these functions.Interface circuit 73 is connected to the control and the state part 21 of system bus 13.The order of self-controller in the future is provided to state machine 75, and state machine 75 then provides the specific control of other circuit so that carry out these orders.Control line 77-81 connect state machine 75 and said other circuit, as shown in Figure 2.Status information from state machine 75 is sent to interface 73 through line 83, to be used for being transferred to controller 11 through bus portion 21.

Memory cell array 41 and 43 enable nand gate are preferred at present, but also alternative use other structure of NOR for example.Can be as the instance of the NAND flash memory of the part of accumulator system and operation thereof with reference to the 5th, 570,315,5,774,397,6; 046,935,6,373,746,6; 456,528,6,522,580,6; 771,536 and 6,781, No. 877 United States Patent (USP)s and No. 2003/0147278 open case of U.S. Patent application.

By the circuit diagram of Fig. 3 exemplary NAND array is described, it is the part of memory cell array 41 of the accumulator system of Fig. 2.A large amount of global bit line are provided, but for the purpose of explaining for the sake of simplicity, in Fig. 2, only show four this type of line 91-94.Many memory cell strings 97-104 that are connected in series are connected between one of these bit lines and the reference potential.Use memory cell strings 99 as representative, a plurality of charge storage memory cell 107-110 are connected in series with selecting transistor 111 and 112 at arbitrary end place of string.When making the selection transistor conduction of a string, said polyphone is connected between its bit line and the reference potential.Then a memory cell in the said string is carried out one-time programming or read.

Individually extend on the charge storage cell of the word line 115-118 of Fig. 3 memory cell in each string in many memory cell strings, and the transistorized state of selection at each end place of grid 119 and 120 control strings.Make the memory cell strings of sharing common word line and control gate line 115-120 form the block of memory cells 123 that to wipe jointly.This element block contains the unit of minimal amount that can disposable physical erase.Once to programming along the row of memory cells of one of word line 115-118.Usually, with predesigned order the row of NAND array being programmed, is with near the row beginning of the word line 118 of the end of the string that is connected to ground connection or another common potential in the case.Then, so in whole block 123, carry out programming along the column of memory cells of word line 117.At last to programming along the row of word line 115.

Second block 125 is similarly, its memory cell be series-connected to first block 123 in the identical global bit line that string connected, but have different one group of word line and control gate line.By line control circuit 55 word line and control gate line are driven into its suitable operating voltage.If have an above plane or subarray in the system, the plane 1 and 2 of Fig. 2 for example, the common word line that so a kind of memory construction use is extended between it.Alternatively can there be two above planes or the subarray of sharing common word line.In other memory construction, the word line of indivedual planes of individual drive or subarray.

Described in the open application case of quoting like some NAND patents and preceding text, but the operational store system with the detectable charge level more than two of storage in each charge storage cell or zone, and then in each above data bit of storage.The charge storage cell of memory cell is the conductive floating gate utmost point the most normally, but alternatively can be non-conductive dielectric charge material for trapping, discloses described in the case like No. 2003/0109093 U.S. Patent application.

The tissue of the flash memory cell array 7 (Fig. 1) of the instance during Fig. 4 conceptual illustration further describes as hereinafter.Four planes or the subarray 131-134 of memory cell can be on single integrated memory unit chips, on (two planes on each chip) on two chips or on four independent chips.Arrange that specifically not heavy b wants concerning the argumentation of hereinafter.Certainly, in system, can there be the plane of other number, for example 1,2,8,16 or more.The plane individually is divided into the block of memory cells of showing with rectangle among Fig. 4, for example is arranged in the block 137,138,139 and 140 of plane 131-134 separately.Can there be dozens of or hundreds of blocks in each plane.As stated, block of memory cells is the unit of wiping, and is the minimum number destination memory location that physically can wipe jointly.Yet, be that unit comes the operation block with bigger first block for the concurrency that increases.To be linked in logic together to form first block from a block on each plane.Show that four block 137-140 form a first block 141.All unit in unit's block are wiped usually jointly.The block that is used to form first block need not to be limited to its identical relative position in plane separately, shown in second yuan of block 143 being made up of block 145-148.Although preferably on all planes, extend first block, can in Different Plane, dynamically form ability and come the operational store system with any or whole first block in one, two or three blocks to obtain high system performance.This allows the size of first block more closely with in the one-time programming operation to can be used for the data quantity stored coupling.

In order to operate purpose, again individual block is divided into page of memory cells, like Fig. 5 explanation.For example the memory cell of each respectively is divided into eight page or leaf P0-P7 among the block 131-134.Perhaps, can there be 16,32 or multi-memory page of cells more in each block.Page or leaf is a data programing and the unit that reads in block, contains the minimum data amount of one-time programming.In the enable nand gate of Fig. 3, page or leaf is formed by the memory cell of word line in block.Yet, can the said page or leaf in two or more blocks be linked into metapage in logic in order to increase accumulator system operation concurrency.Explanation metapage 151 among Fig. 5, it is by forming from the Physical Page of each among four block 131-134.Metapage 151 for example comprises in said four blocks the page or leaf P2 in each, but the page or leaf of metapage need not in each of block, necessarily to have identical relative position.Although preferably on all four planes multiple programming with read maximum amount of data obtaining high system performance, but also the operational store system to form any or the whole metapages that has in, two or three pages or leaves in the independent block in Different Plane.This allow programming and read operation adaptively with the data volume coupling of parallel processing easily, and the part maintenance that has reduced metapage is not with the chance of data programing.

Be used for using most memory management techniques of independent page or leaf and independent block management data can be applied to metapage and first block equally at every turn.Similarly, use the technology of metapage and first block also to can be applicable to independent block and independent page or leaf usually.Substantially, the instance that uses page or leaf and block to provide will be interpreted as and can be applicable to metapage and first block.Similarly, the instance that provides with respect to metapage and first block will be interpreted as and can be applicable to page or leaf and block usually.

Illustrated like Fig. 5, the metapage that is formed by the Physical Page on a plurality of planes contains along the memory cell of the word line row on those a plurality of planes.Be not to be programmed in all unit in the word line row simultaneously, but more common it being replaced with two or more staggered groups programmed each group's storage data page (in single block) or data metapage (on a plurality of blocks).Through the memory cell that replaces is programmed at every turn, need not all provides the peripheral circuit unit that comprises data register and sensor amplifier to each bit line, but between contiguous bit line the time share.This has saved the required substrate space amount of peripheral circuit, and allows to encapsulate memory cell with the density that increases along row.In addition, preferably to programming simultaneously, so that the concurrency that maximization can obtain from given accumulator system along each unit of row.

Referring to Fig. 3, through at least one end along NAND string provide two row selecting transistors (not shown) rather than shown in single row, come most convenient ground accomplish with data be programmed into simultaneously along delegation whenever in a memory cell.The selection transistor of delegation then is series-connected to its separately bit line with every in the block at a distance from one in response to a control signal, and selection transient response of another row whenever is series-connected to its bit line separately at a distance from one in what another control signal will get involved.Therefore two data pages are written in every row of memory cells.

Data volume in each logical page (LPAGE) is one or more individual data sector of integer normally, contain the data of 512 bytes on each sector convention.Fig. 6 shows two sectors 153 of page or leaf or first page data and 155 logical data page or leaf.Each sector contain usually to some extent storage user or system data 512 bytes part 157 and be used for relating to the data of part 157 or relate to the Physical Page that wherein stores data or the byte 159 of another number of the overhead data of block.The byte number of overhead data is generally 16 bytes, thereby for each is 528 bytes altogether in sector 153 and 155.Overhead part 159 can contain that the ECC, its logical address, the block that during programming, calculate from data division 157 have been wiped and the experience of the number of times of programming again counting, one or more control flag, operating voltage level and/or analog, adds the ECC that overhead data 159 from then on calculates.Perhaps, overhead data 159 or its part can be stored in the not same page in other block.Along with the concurrency increase of storer, therefore the data storage capacity increase of first block and the size of data page and metapage also increase.Data page then can contain two above data sector.Because two sectors and two data pages of each metapage in the data page, so there are four sectors in a metapage.Therefore each metapage stores the data of 2048 bytes.This is the concurrency of height, and further increases when number memory cells increases in can even being expert at.From this reason, extend the width of flash memory so that increase the data volume in page or leaf and the metapage.

Physically less Reprogrammable non-volatile memory card and flash drive that preceding text are pointed out are commercially available, and it has the data storage capacity of 512 megabyte (MB), GB (GB), 2GB and 4GB, and it is higher to become.Fig. 7 explains the main frame modal interface between the mass storage system therewith.Host process is by the application software of main frame execution or the data file of firmware program generation or use.The word processing data file is an instance, and the plotting file of computer-aided design (CAD) (CAD) software is another instance, for example mainly is present in the general main frame such as PC, laptop computer.The document of pdf form also is this file.The picture digital camera is that the every width of cloth figure that is stored on the memory cards produces a data files.Cellular phone is used to the data of the file on the internal storage card, for example telephone directory.The PDA storage is also used some different files, for example address file, Cal. Files etc.In any this application, memory cards also can contain the software of operating host.

Accumulator system especially is implemented in the accumulator system in the removable card, can be via standard interface and different main-machine communications.Different main frames can use different interfaces to communicate by letter with accumulator system.Two class interfaces are to use the interface and the interface that uses based on the addressing system of file of the logic addressing system with shared logical address space.

The LBA interface

Fig. 7 explains the logic interfacing commonly used between main frame and the accumulator system.Continuous logical address space 161 enough thinks that greatly all data that can be stored in the accumulator system provide the address.Usually be the increment of data clustering with the host address spatial division.Each is clustered in the given host computer system can be greatly typical between 4 and 64 sectors through being designed to contain many data sector.The data of 512 bytes are contained in the standard sector.

Show in the instance of Fig. 7 and produced three files 1,2 and 3.The application program of on host computer system, moving produces each file as one group of ordered data, and through unique name or other reference it is discerned.Still the unallocated enough available logical address space of other file of giving is assigned to file 1 by main frame.File 1 is shown as and is assigned continuous utilogic address realm.Usually also from specific purpose and distribute address realm, for example, then avoid these particular ranges, even do not utilize these addresses during to data assignment logic address yet at main frame for the storage data to the particular range of host service function software.

When main frame produced file 2 after a while, main frame was assigned the continuation address of two different range similarly in logical address space 161, as shown in Figure 7.File need not to be assigned continuous logical address, but can be the address fragment between the address realm of distributing to other file.The another file 3 that this instance is then showed main frame and produced is assigned with the host address space the previous unallocated file 1 and 2 and other parts of other data of giving.In this example, file 1, file 2 and file 3 are all assigned the part of giving shared logical address space (logical address space 161).

Main frame comes the trace memory logical address space through keeping file allocation table (FAT), wherein assigns the logical address to each Hosts file to be maintained by main frame.The FAT table is stored in the nonvolatile memory and in the mainframe memory usually, and along with new file is stored, other file is deleted, file is modified or the like, the FAT table is upgraded by main frame continually.For example when Hosts file was deleted, main frame then came the logical address releasing of before distributing to deleted file is distributed through upgrading the FAT table, can be used for now using with other data file to show them.Therefore the logical address of using among the FAT can be described as logical block addresses (LBA), uses at the interface that to the data from different files is this logic of class addressing on the shared logical address space and can be described as the LBA interface.The agreement of the communication of the logical address of the data that similarly, are used for transmitting to use can be considered the LBA agreement.

Main frame is indifferent to accumulator system controller wherein and is selected to be used for the physical location of storage file.Typical main frame only knows that its logical address space and its distributed to the logical address of its each file.On the other hand, accumulator system is through typical main frame/card, only knows the part that the data of logical address space have been written to, but do not know to distribute to the logical address of particular host file, or even do not know the number of Hosts file.The accumulator system controller is used to store or the logical address of retrieve data converts the unique physical address in the flash memory cell array that wherein stores host data to main frame provided.Block 163 these logics of expression are to the worksheet of physical address translations, and it is kept by the accumulator system controller.

The accumulator system controller through programming so that so that the performance of system maintain in the block and first block that high-grade mode is stored in data file memory array 165.In this explanation, use four planes or subarray.The maximum concurrency programming and the reading of data that on the whole first block that forms by block, preferably allow with system from each plane.At least one first block 167 is stored by the operation firmware of Memory Controller use and the reservation block of data through distributing as being used to usually.Can distribute another yuan block 169 or a plurality of first block to be used for storage host function software, main frame FAT table etc.The major part that keeps amount of physical memory is to be used for storing data files.Yet Memory Controller does not know how main frame distributes the data that received between its each file object.Memory Controller from host interaction the full content known usually be exactly be written to specific logic address by main frame data storage in the respective physical address of keeping to physical address table 163 by the logic of controller.

In typical accumulator system, the memory capacity that manys several blocks than essential is provided, with the amount of the data in the memory address space 161.Can provide one or more conducts in these additional blocks to be used to replace to become the redundant block of defective other block in the storer length of life.The logic groups of the block that the individual metablocks block contains can change because of a variety of causes usually, comprises redundant block is given the defectiveness block of first block to original appointment replacement.One or more additional blocks of for example first block 171 maintain usually in the block pond of wiping (block pool).When main frame writes data into accumulator system, controller will be converted to the physical address in the first block in the erase blocks pond by the logical address that main frame is assigned.Be not used for other yuan block of data storage in logical address space 161 then through wipe and specify as through the pond block wiped for during the follow-up data write operation, using.

When the data of original storage become out-of-date, frequently override the data that are stored in the particular host logical address by new data.As response, the accumulator system controller writes the block through wiping with new data, and then changes logic to the new physical blocks of physical address table with these those logical address place data of identification storage to those logical addresses.Then wipe the block that contains those logical address place raw data and make it can be used for the storage of new data.If when writing beginning, in the preparatory erase blocks from the erase blocks pond, do not have enough memory capacity, so this usually wiping must be taken place before the current data write operation can be accomplished.This possibly influence the system data program speed unfriendly.Memory Controller is only recognized usually makes the data at given logical address place become out-of-date by main frame at main frame when its same logical address writes new data.Therefore many memory blocks possibly stored this type of invalid data and last certain hour.

The size of block and first block increases gradually, so that use the integrated circuit memory area of chip effectively.This causes the individual data of larger proportion to write the memory capacity of data quantity stored less than first block, and in many cases even less than the memory capacity of block.Because the accumulator system controller is directed to the pond unit block through wiping with new data usually, so this can cause the several portions of first block not to be filled.If new data is to being stored in certain updating data in another yuan block, so also expect all the other active data metapages to be copied to new first block from said another yuan block that has with the continuous logical address of the logical address of new data metapage with logical address order.Old first block can keep other active data metapage.This caused the data of some metapage of individual metablocks block to become along with past time out-of-date and invalid, and replaced by the new data with same logical address that is written to different first blocks.

On whole logical address space 161, store data in order to keep enough physical memory space, these data are carried out periodic compression or merging (refuse collection).Also hope to keep the data sector in first block, because this makes that the data that read in the continuous logic address are more effective with the same order of actual as many logical address with it.Therefore carry out data compression and refuse collection with this extra purpose usually.Aspect some of when the receiving unit block data upgrades, storer being managed and the use of first block be described in the 6th, 763, in No. 424 United States Patent (USP)s.

Data compression is usually directed to read all valid data metapages and be written into new block from first block, has wherein ignored the metapage that has invalid data during the course.Also preferably with the metapage with valid data in the physical address order of logical address order coupling of the data of storing arrange said metapage.The number of the metapage that occupies in new first block will not copy to new first block because will not contain the metapage of invalid data less than the number of the metapage that occupies in old first block.Then wipe old block and make it can be used for storing new data.Then can be used for storing other data through merging the extra metapage that obtains with capacity.

During refuse collection, collect the metapage of valid data from two or more yuan block, and it is rewritten in another yuan block, normally a first block in the erase blocks pond with continuous or approximately continuous logical address.When original two or more yuan block has duplicated all valid data metapages, can be with said first block erase for using future.

Data merge with the refuse collection spended time and possibly influence the performance of accumulator system, especially merge or refuse collection need can carried out from the order of main frame under the situation about carrying out before in data.This generic operation usually by accumulator system controller scheduling carrying out on the backstage as far as possible, but the needs of carrying out these operations possibly cause controller to provide busy status signal to main frame, till this operation is accomplished.The instance that the execution of Host Command can be postponed is in the erase blocks pond, not have enough first blocks of wiping in advance to come storage host to hope to be written to all data in the storer; And at first need data merging or refuse collection to remove one or more valid data unit blocks, said first block then can be wiped free of.Therefore concentrate on the control of diode-capacitor storage so that minimize this type of interruption.In following patent application, describe many these type of technology: the title of application on Dec 30th, 2003 is the 107/749th, No. 831 patent application of " Management of Non-Volatile MemorySystems Having Large Erase Blocks "; On Dec 30th, 2003, the title of application was the 10/750th, No. 155 patent application of " Non-Volatile Memory and Method with Block Management System "; On August 13rd, 2004, the title of application was the 10/917th, No. 888 patent application of " Non-Volatile Memory andMethod with Memory Planes Alignment "; The 10/917th, No. 867 patent application of application on August 13rd, 2004; On August 13rd, 2004, the title of application was the 10/917th, No. 889 patent application of " Non-VolatileMemory and Method with Phased Program Failure Handling "; On August 13rd, 2004, the title of application was the 10/917th, No. 725 patent application (hereinafter is referred to as " LBA patent application case ") of " Non-Volatile Memory and Method with Control DataManagement ".

A challenge effectively controlling the operation of the memory array with very big erase blocks is that the data sector number of storing during the given write operation is also aimed at the capacity and the boundary matching of memory block.A kind of method is that the maximum number with the necessary block of a certain amount of data that is less than the amount of filling whole first block less than storage disposes and is used to store the first block from the new data of main frame.The title of on Dec 30th, 2003 application be the use of having described the first block of self-adaptation in the 10/749th, No. 189 patent application of " Adaptive Metablocks ".The 10/841st of application on May 7th, 2004; The title of No. 118 and on Dec 16th, 2004 application is to have described the cooperation on border between the block and the physical boundary between first block in the 11/016th, No. 271 patent application case of " Data Run Programming ".

Storer control also can use the data of the FAT table of free host stores in nonvolatile memory, with operational store system more effectively.A kind of this type of use is to understand main frame when to distribute data identification be out-of-date through the logical address of data is removed.Know that this situation allows Memory Controller to write new data and dispatched wiping of the block that contains this invalid data will know situation usually before to those logical addresses through main frame at it.This is described in 10/897th, No. 049 patent application of the title of application on July 21st, 2004 for " Method and Apparatus for Maintaining Data on Non-Volatile MemorySystems ".Whether other technology comprises monitor host and writes the pattern of new data to storer, be single file so that infer given write operation, or infers when a plurality of file where the border between the file is positioned at.The title of application on Dec 23rd, 2004 is that the 11/022nd, No. 369 patent application of " FAT Analysis for Optimized SequentialCluster Management " described this type of technological use.

For the valid function accumulator system, controller need be use up it and understood possibly by main frame and assign the logical address to the data of its respective files.Then controller can be stored in data file in single first block or the first block group, rather than when the unknown of file border, is dispersed between first block of greater number.The result is that data merge and the number of times and the complicacy of garbage collection operations reduce.So improvement in performance of accumulator system.But when main frame/memory interface comprised aforesaid logical address space 161 (Fig. 7), memory controller was difficult to understand the host data file structure morely.

Referring to Fig. 8, typical logical address main frame/memory interface of having showed of key diagram 7 by different way.By the data file assign logical addresses of main frame to the main frame generation.Accumulator system is then seen these logical addresses and it is mapped in the physical address of actual storage memory of data blocks of cells.

File interface

The interface based on file that main frame and being used to is stored between the accumulator system of mass data does not use logical address space.Main frame changes each file of offset address addressing logically through the data unit (for example byte) in unique file ID (or other unique reference) and the said file into.This file address is directly to give the accumulator system controller, and the accumulator system controller is then preserved its oneself the data about each Hosts file and physically is stored in table where.Availablely implement this new interface with respect to Fig. 2 to the described same accumulator system of Fig. 6 like preceding text.Is the mode that accumulator system is communicated by letter with host computer system with Fig. 2 to the main difference part of content described in Fig. 6.

Supporting paper interface among Fig. 9, it should be compared with the LBA interface of Fig. 7.The side-play amount of the data in the identification of each in the file 1,2 and 3 and the file of Fig. 9 is directly transferred to Memory Controller.This logical address then converts the first block of storer 165 and the physical address of metapage to by Memory Controller function 173.

Figure 10 is the supporting paper interface also, and it should be compared with the logical address interface of Fig. 8.The FAT table that the logical address space of Fig. 8 and main frame are kept does not exist in Figure 10.But discern the data file that produces by main frame to accumulator system through the side-play amount of the data in reference number of a document and the said file.Accumulator system is then directly with the physical blocks of said File mapping to memory cell array.

Because accumulator system is known the position of the data that constitute each file, therefore can after host deletes file, wipe these data immediately.This is impossible for typical logical address interface.In addition, discern host data through file rather than use logical address, the accumulator system controller can reduce the mode of the needs that carry out frequent data item merging and refuse collection and store data.Therefore the frequency of data replicate run significantly reduces with the data volume of being duplicated, and then increases the data programing and the reading performance of accumulator system.

Instance based on the interface of file comprises the interface that uses the immediate data file storage.With Alan W.Sinclair people name or with Peter J.Smith in the 11/060th of application on February 16th, 2005; 174,11/060; 248 and 11/060; The title of No. 249 patent application co-pending and Alan W.Sinclair and Barry Wright application is to have described immediate data file storage memories system in the 60/705th, No. 388 provisional application case (hereinafter being referred to as " immediate data file storage application case ") of " Direct Data File Storage in Flash Memories ".

Because the immediate data file interface of these immediate data file storage application cases of Fig. 9 and Figure 10 explanation is simpler than the above-mentioned logical address space interface of Fig. 7 and Fig. 8 explanation; And allow accumulator system to carry out better, so the immediate data file storage is preferred to many application.But host computer system mainly is configured to LBA interface operation at present, and the accumulator system and the most main frame that therefore have the immediate data file interface are incompatible.Therefore need the ability with any interface operation be provided to accumulator system.

Double nip

For some accumulator systems, especially be included in can with different main frames those accumulator systems in the removable memory cards that connects that are situated between, back compatible property is a major issue.Many host computer system uses and Fig. 7 and shown in Figure 8 similarly based on the form of the storage of sector, and in these host computer systems some possibly can't easily be suitable for operating with for example Fig. 9 and the storage based on file shown in Figure 10.Therefore, need have can with the accumulator system of using logic addressing interface or being situated between and connecing based on the main frame of the interface of file.The main frame that is inserted in the LBA interface and can allows to use the logic addressing based on the LBA protocol adaptor between the rear end of file with data storage in the memory array that data are managed as file.

The title of on August 3rd, 2005 application be the 11/196th, No. 869 patent application of " Interfacing systems operating through a logical addressspace and on a direct data file basis " described make accumulator system can with the system that uses logic addressing interface or connect based on main frame Jie of the interface of file.Figure 11 explains this system.This instance has made up the storage operation based on file of host service function and Fig. 9 of Fig. 7 in accumulator system, add the address translation 172 of an interpolation.Address translation 172 is mapped to the logical address group on the storage space 161 in indivedual logical file a-j of space, modified address 161 ' upward displaying.Preferably whole logical address space 161 is divided into these logical files, so the number of logical file depends on the size of logical address space and the size of indivedual logical files.Each logical file all contains the data of the group of continuous logic address on 161 of having living space.Preferably make the data volume in each logical file identical, and said amount equal the data storage capacity of a first block in the storer 165.The unequal size of logical file and/or with the different size of memory capacity of the block of storer or first block that yes is possible, but be not preferred.

Each interior data is represented by the logic offset address in the file among the respective files a-j.The file identifier of logical file and the data offset physical address in 173 places convert storer 165 into.Logical file a-j directly is stored in the storer 165 through identical process and the agreement of describing in the immediate data file storage application case.Said process is identical with the process that is used at the data file 1-3 of storer 165 storage maps 9; The given data amount that different is in each logical file makes this process be easier to, and especially equals under the situation of capacity of block or first block of storer in said amount.Show among Figure 11 among the logical file a-j each is mapped to the not same person in first block of storer 165.Also expect based on the data storage of file identical or equivalent mode and the host interaction of current logical address accumulator system to connect with Jie through design with main frame.Through indivedual logical files being mapped in the corresponding individual memory unit block, performance and temporal characteristics identical when using the logical address space interface have been realized with immediate data file interface accumulator system in essence.

Figure 12 explains the method for Figure 11 by different way.Figure 12 is identical with the logical address accumulator system operation of Fig. 8, but has the function that logical address space is divided into the increase of logical file (step 172 of the Figure 11 that has just described).In addition, Figure 12 " file data being mapped to the table of physical store block " substituted " logical address being mapped to the table of physical store block " of Fig. 8.Logical address can be considered at the interface that uses the LBA system and based on the part of the LBA protocol adaptor between the rear end of file to logical file conversion 172.

The rear end storage system based on data file that warp designs with the Figure 11-12 that works with main frame through the traditional logic address space interface also can be added with the immediate data file interface.All translate to the storage element block address from the host data file of file interface with from the logical file of logic interfacing.Then through carry out the immediate data file protocol with data storage in those addresses of storer.This agreement comprises the immediate data file storage technology of the immediate data file storage application case of before having listed.

Through the pocket memory card with two types of HPIs or the removable accumulator system of flash drive or other form are provided; Storer can use with the most recent main frame with the logical address space interface operation; Use with the main frame of receiving storer that its file directly is situated between, or can between two types of main frames, exchange.This user who allows to have the main frame of newer interface based on file uses storer with its effective and efficient manner, has the back compatible property with the traditional logic address space interface simultaneously.And, owing to identical logical file is one to one realized identical in essence performance and temporal characteristics to the mapping of first block.Storer with two HPIs allows the user to obtain the newer immediate data file interface of said storer, the substrate that its expansion that still can be used for having the main frame of traditional logic address space interface is simultaneously installed.It provides the method from current interface shift to the immediate data file interface.

Figure 13 explanation has the accumulator system 300 of two HPIs.The host data file (HF1, HF2...HFn) that accumulator system storage is directly supplied through file interface 307 by main frame and from the logical file of changing by LBA protocol adaptor 301 (LFa, LFb...LFm) of LBA interface 305.Rear end 303 based on file does not need discriminate between logical file and Hosts file, but preferred through optimizing the file to handle two types.So; Logical file is equal to the logical. group of the system of describing in the LBA patent application case; Therefore and from the angle of HPI, the performance of accumulator system 300 is complementary with the performance of the system with logical address space interface described in the LBA patent application case.

Except host data is converted into the logical file form from the logic addressing format, the specific LBA order that LBA protocol adaptor 301 can receive in response to specified conditions or from main frame and produce with based on the compatible order in the rear end of file 303.The instance that produces the order that is used for immediate data file rear end in response to these a little conditions is showed in following table.

Condition	The immediate data file command that is produced
		The beginning of LBA order	Write _ pointer this order of < file ID>< side-play amount>offset address in the current file is set, next write command will be in this place, address operation.
Current file is not opened and is opened _ file=maximum	Guan Bi<Wen JianID>+ idle+time-out<Up to _ or not busy>This order group closes the file of access least recently, and impels device to carry out all garbage collection operations to file co-pending.
		Current file is not opened	Da Kai<Wen JianID>+ write _ Zhi Zhen<Wen JianID><Pian Yiliang>This order group opens current file, and offset address is set, and next write command will be worked at this place.
Current file changes	Xie Ru<Wen JianID>+ flow+write _ Zhi Zhen<Wen JianID><Pian Yiliang>This order group writes the data of the last file of accumulation, and the offset address in the current file is set, and next write command will be in this place's operation.
		Program area block is full	Xie Ru<Wen JianID>This order group of+stream writes the data of the current file of accumulation.
Program area block has been expired and effective _ page or leaf>Minimum value	Guan Bi<Wen JianID>+ idle+time-out<Up to _ or not busy>This order group closes current file, and impels device to carry out all garbage collection operations to file co-pending.
		The end of LBA order	Xie Ru<Wen JianID>+ stream+preservation _ buffering<Wen JianID>This order group writes the data of the current file of accumulation, and impels the not programming data that is retained in the impact damper to be programmed into " exchange block " in the flash memory.

Object interface

Various interfaces based on file are used in and transmit data between the electronic installation.Some agreements provide the file of the indication with predetermined size and said size.The size of file remains unchanged in this system usually, makes this system possibly be inappropriate for the application that need edit file.Yet for transmitting file from a device to another device, said agreement maybe be favourable, and can allow high-grade security.The indication of size was sent before sending file data usually.The agreement of file that transmission has the predetermined size of file size designator can be considered the object agreement.The picture transportation protocol (PTP) that a kind of object agreement is a Microsoft.Another kind of this agreement also is the media transfer protocol of Microsoft, is also referred to as media transmission protocol (MTP).The object agreement is particularly suited for sending the file that contains predetermined amount of data, for example digital photos or MP3 music file.For instance, said agreement is used between digital camera and the PC and transmits digital photos, or the MP3 music file is sent to the MP3 player from PC.

Media transfer protocol (MTP) provides support and is scheduled to the object interface that big or small file object transmits.Its fundamental purpose is the communication between the device that allows temporarily to link together, and wherein each device has very big memory capacity.Interface allows the exchange of binary data object between the device, and device connects the enumerating of content of another device.Hereinafter is listed some characteristic of MTP interface.Yet other object interface can have different characteristic.Title from Microsoft provides the description more completely to MTP in the document of " Media Transfer Protocol Enhanced ".

1. communication protocol

1.1 initiator and responsor: only between two devices, exchange, one of them device serves as initiator and another device serves as responsor at every turn.Initiator is initial through transmit operation and the device of the action of responsor.The not initial any action of responsor, and the operation that initiator sends sent response.The content of responding device and the operating process in the control protocol should enumerated and understand to the device that serves as initiator.The device that initiator is normally more powerful than responsor.The instance of responsor is simple content generation device (for example digital camera) and simple content output apparatus (for example portable audio player).

1.2 session: session is wherein between initiator and responsor, to keep the communications status that is connected.Session provides the context in order to references object, and guarantees that the responder device state can not change under the situation of warning not sending to initiator.Session is opened by initiator, and is closed by initiator or responsor.Device can be kept a plurality of sessions of opening simultaneously.Initiator is assigned unique identifier to session when session is opened for the first time, and when transmit operation, uses identifier to discern session.

1.3 affairs: any action that from initiator, occurs is carried out in affairs, and affairs are the standard sequence of stages that input parameter, binary data exchange are provided and have the response of parameter to action invocation mechanism.Data stream in each stage is unidirectional.The operation initial during, data only flow to responsor from initiator.At the response duration to institute's solicit operation, data only flow to initiator from responsor.During the binary data switching phase, data can flow on arbitrary direction, but never on both direction, flow.Must carry out two-way binary data exchange through a plurality of operations.Initiator is assigned an identifier to each affairs.To assign to first initial in the session operation through the identifier of definition, and identifier increased progressively 1 to each continuous affairs.

Affairs are formed by reaching three phases: operation requests stage, data phase and response phase.The operation requests stage is shared identical identifier with response phase, and optional data phase is present between two other stage when needed.

The operation requests stage is by the transmission of operation requests data set (operation that its identification is called by initiator), wherein its context that should be performed (session and affairs) and one group of limited parameter are formed.

Optional data phase is followed the operation requests stage.There is the operation decision that is defined in the stage by operation requests in it.Data can be the transparent data collection that in agreement, defines, or can be through exchange to be stored in the binary data on the receiving trap.Actual data transmission can relate to Container Format transmission data in the data phase, or data are divided into several bags, and is required like specific transmission just in use.

In response phase, the fixed data collection is transferred to initiator with the information of report about previous affairs, for example success/failure from responsor.

1.4 incident: incident is mainly sent as taking the photograph the mode that mode is transmitted information or alarm in the past by responsor.Be different from operation, need do not reply or move incident.Incident need or be operated the affairs asynchronous communication with data transmission.Transmission should define the process that incident can be interlocked with data stream during affairs.

1.5 synchronous and asynchronous transaction: all affairs in the communication protocol all are synchronous, promptly can't initial new operation before last operation is accomplished fully.Can be through the incident of sending by responsor when carrying out in the backstage on operating in device, initial through operation is divided into (it makes the operation beginning) and process are kept watch on and are simulated asynchronous operation.If when asynchronous operation is in processing, attempt new operation, responsor responds with busy status of fail so, and initiator should be attempted after a while once more.

2. message data set

When swap data was collected, it was collected with the predefine structure that is called data set.Have three kinds of message data sets, it can come access through using suitable operation.

2.1 device information data set: the device information data set provides the description to device, and major part is static.

2.2 object information data set: the object information data set provides the summary to the core attribute of object.These core attribute comprise the size of the data component of object.Attribute also comprises relevance, and it can be used for making associated with data objects, and the hierarchical file system on the tracing device.Can not rely on any paths or the UNC that are exclusively used in specific file system and come the file classification on the indication device.The attribute of object also can be concentrated retrieval at object attribute data.

2.3 storage device information data set: the memory storage that comprises in the storage device information data set tracing device.Description comprises max cap. and still has free space to be written, and can comprise file-naming convention or bibliographic structure convention in the use.

3. attribute

3.1 device attribute: the setting or the status condition of device property recognition means, and be not linked to any data object on the device.The device attribute can be read-only or read-write.Attribute kit is contained in device attribute description data centralization, and can be provided with or retrieve through using suitable operation.

3.2 object properties: object properties are provided for exchanging the mechanism with the separated object descriptive metadata of object own.The principal benefits of object properties is to allow the enumerating fast of big memory storage, no matter and file system how.Object properties provide the information about the object on the device, and the value of specifying it to comprise.Attribute kit is contained in object attribute data and concentrates, and can be provided with or retrieve through using suitable operation.

4. object handle

Object handle provides the identifier to the unanimity reference (being unique) of the object logic on the device in session.The value of object handle does not have specific meaning.Responsor is in response to opening session operation and be that object in the device produces the object handle array from initiator.Initiator is operated by means of the acquisition object handle and is obtained object handle, and this impels responsor to send the object handle array to initiator.Operate when defining object to be sent when initiator use to send object information, responder device is distributed object handle, and in the response phase of operation, it is turned back to initiator.When session was closed, all object handles all lost efficacy, and must be obtained again by initiator.Device can keep identical object handle maybe can change object handle to be used for next session.

5. object reference

Because object interface is the agreement that is independent of file system, so can't between object, form complicated link through embedding filename.Defined an abstract refer-mechanisms to allow object reference arbitrarily.Reference is unidirectional, and can't confirm the object which object reference is given under the situation of all references of all objects in not checking device.Can be provided with or retrieving reference through using suitable operation.The object of file handle reference must be consistent between session.The reference of object to deletion can not be by error with reference to another object.Any object handle should not reuse, and perhaps device should be deleted all references to an object together with object.

6. operation and response

Operation is defined in communicating by letter of taking place between the inherent initiator of affairs and the responsor.Responsor can be used as parameter and on operation requests, transmits the startup required information that works.Can send 5 parameters.Extra information also can be with the predefined data sets transmission in the data phase of affairs.After each operation, responsor returns has response data sets that reaches 5 parameters and the answer code of indicating operating result.Specify a large amount of operations.The instance of the sequence of operation is " transmission object information " operation of following " transmission object " operation; Initiator sends object through this operation to responsor; And " acquisition object information " operation of following " acquisition object " operation, initiator receives the object from responsor through this operation.

Figure 14 A shows the instance of the affairs between initiator 410 and the responsor 412.Initiator 410 can be PC, and responsor 412 can be MP3 music player or digital camera.Initiator 410 and responsor 412 for example communicates through USB cable through being connected allowing.At first, in the stage, initiator 410 is discerned the operation of calling as " transmission object information " operation in operation requests.The second, in data phase, initiator 410 sends object information 414 to responsor 412.Object information 414 is about being sent to the information of the special object of responsor 412 by initiator 410.Object information 414 was sent by initiator 410 before its referent is sent out.Object information can comprise the multiple information about object, comprises the size of object.Will send to as if the situation of MP3 music file under, the size of MP3 music file is sent as the part of object information.The 3rd, in response phase (after object information 414 is received by responsor 412), responsor 412 indications have received object information 414.Affairs can finish at this moment.。

Figure 14 B shows the affairs follow Figure 14 A and as second affairs of the part of the same session between initiator 410 and the responsor 412.Second affairs are " transmission object " affairs.At first, in the stage, initiator 410 identifications are as the operation of " transmission object " operation calls in operation requests.The second, in data phase, initiator sends object 416 to responsor 412.Object 416 can be the MP3 music file or has the for example digital photos of JPEG, GIF or bitmap format.The object information 414 of object 416 (include file size) sends to responsor 412 by initiator 410, as said with respect to Figure 14 A.The 3rd, in response phase (after receiving object 416), responsor 412 has received object 416 to initiator 410 indications.Affairs can finish at this moment.

The object protocol adaptor

In an embodiment of the present invention, the object protocol adaptor is provided in accumulator system, its main frame that allow to use the object agreement with use as stated accumulator system based on the rear end of file to be situated between to connect.The object protocol adaptor receives data and order according to the object agreement through object interface, and before sending data to the back-end and ordering, carries out suitable translating.In an example, main frame uses MTP to come to connect with accumulator system Jie, and brings in the storage data behind the accumulator system use immediate data file.The object protocol adaptor is carried out suitable conversion to order between HPI and rear end and data.

Figure 15 shows and to be inserted in object interface 520 and based on the instance of the object protocol adaptor between the rear end 522 of file.Object interface 520 is the interfaces that are used for the file object (for example MTP or PTP object) of predetermined size.The size of object was sent before object is sent by main frame.Object protocol adaptor 524 management agreements are at said agreement internal storage system 526 and main-machine communication.Object protocol adaptor 524 is also managed and is used for the affairs of the message exchange of main frame and carries out affairs.Affairs are according to carrying out between object protocol adaptor and the rear end based on file based on the agreement of file.

Object protocol adaptor 524 when new file object is sent (according to the object agreement) by main frame through sending the opening and closing that appropriate command are come management document to rear end 522 based on file.In particular, because file has predetermined size in the object agreement, so object protocol adaptor 524 is responsible for close file when receiving the file data of scheduled volume.Use the main frame of object agreement will not send independent end of file indicator usually, so object protocol adaptor 524 produce end of file indicator and send it to the rear end 522 based on file when receiving whole file object.In this way; When accumulator system 526 receives a plurality of file from the main frame that uses the object agreement; Rear end 522 based on file can not keep File Open, till it reaches opening file of certain maximum number (as taking place usually in the immediate data file rear end).Alternatively, each file receives as complete object, and after said complete file is received, receives the order from object protocol adaptor 524 based on the rear end of file, feasible rear end close file based on file.This has reduced keeps many burdens that open file.But in case closing of a file with regard to schedule file to carry out refuse collection.

But the state of object protocol adaptor diode-capacitor storage system.In particular, for the accumulator system with immediate data file rear end, define three kinds of states, it allows accumulator system to change its operation in response to Host Command.These three kinds of states are: " leaving unused ", " subsequent use " and " closing ".Corresponding state order in response to from main frame comes initial these three kinds of states, and wherein main frame uses immediate data file command collection or equivalent.Use at main frame under the situation of object agreement, the object protocol adaptor can produce status command.If main frame uses the object agreement that comprises said order, the object protocol adaptor can come to the rear end transmit status order based on file in response to the equivalent command that receives from main frame so.Perhaps, the object protocol adaptor can be based on inferring that from other factors the state of main frame produces status command.For instance, main frame can provide certain indication: it will can not remove power in a period of time, and as response, the object protocol adaptor can send " leaving unused " status command to the rear end based on file.Similarly, main frame can indicate it will remove power, or the object protocol adaptor can infer based on the main frame behavior removing power, and as response, the object protocol adaptor can be to sending " closing " order based on the rear end of file.

One of major function of object protocol adaptor is that be to be used for stream data based on the rear end of file since the host data conversion (according to the object agreement) of the predefine file of main frame reception future.When the main frame (serving as the initiator among the MTP) that uses the object agreement sends the file with predefine size and require the response from responsor, can not be configured to provide this response usually based on the rear end of file.In particular, under the situation of using immediate data file rear end, data are fluidized usually, and do not have the equivalent of the response phase of MTP.The object protocol adaptor will convert stream data into from the object of main frame, and when receiving all data of object, produce the suitable response to main frame.

In the object agreement of for example MTP, the operation of reading or write object is to be used to transmit the independent operation about the information of said object before.This information transfer operation is that " setting " to object properties (object information) operated under the situation of the object that is just writing.This information transfer operation is " acquisition " operation to object properties under the situation of the object that is just reading.Object properties have the form of the data set that comprises object length.

When object just write the accumulator system 526 of Figure 15, the data set of object was at first received from main frame by object protocol adaptor 524.Main frame is regarded as initiator in the MTP item.The information of data centralization is used to control the subsequent transaction of implementing write operation.Data set also as metadata store in accumulator system 526.During write operation, the data volume that object protocol adaptor 524 countings transmitted from initiator during the data write operation stage.The end of identifying object data during from then on object protocol adaptor 524 is counted.The object protocol adaptor then produces the response that sends to main frame.Object protocol adaptor 524 also sends order with close file to the rear end 522 based on file.

When just from accumulator system 526 reading object of Figure 15, the metadata of object is at first obtained from the rear end 522 based on file by object protocol adaptor 524.The metadata of object is then used in the subsequent transaction that read operation is implemented in control.The metadata that object protocol adaptor 524 also will be stored in the accumulator system 526 sends to main frame as the data set that is used for object.Object protocol adaptor 526 countings are the data volume from transmitting based on the rear end 522 of file during the data read operation stage.The end of recognition data during from then on object protocol adaptor 524 is counted.Object protocol adaptor 524 then produces the response that sends to main frame after object sends.

The object protocol adaptor is at the object information data set that is used for the object agreement of MTP for example and be stored between the metadata based on the accumulator system of file of those accumulator systems of for example using immediate data file rear end and translate." metadata " is to be used in reference to the term of generation about the data of object, and said data and object are stored and management separately discretely.Therefore, the object information data set among the MTP is an instance of metadata.Storing metadata can be come with employed different-format of object agreement or same format in immediate data file rear end.Term " file _ information " also is used for the metadata of immediate data file system.When object protocol adaptor 524 receives the metadata related command from main frame, object protocol adaptor 524 can with order translate to based on the compatible form in the rear end of file 522.Yet in some cases, without any need for translating, because be used for the order and rear end 522 compatibilities of the metadata of object agreement based on file.

Metadata is the information that is associated with file that is produced by main frame.The characteristic of metadata and content are confirmed by main frame, and it usually can be by the device decipher of storage file and metadata.The metadata order is used for initial metadata input and output operation to the specified file of being stored by immediate data file rear end, and the offset address value in the definition metadata.Metadata order can be produced by the object protocol adaptor when the corresponding command that receives from main frame about the MTP data set.Instance by the metadata of immediate data file back-end system use is ordered is showed in following table.

Order	Parameter	Describe
			Metadata _ write	< file ID >	The metadata of specified file is write the offset address that the currency by metadata _ write _ pointer defines
Metadata _ read	< file ID >	Read the metadata of specified file at the offset address place that the currency by metadata _ write _ pointer defines
			Metadata _ write _ pointer	< file ID>< side-play amount >	For the metadata of specified file _ write _ pointer defines new currency
Metadata _ read _ pointer	< file ID>< side-play amount >	For the metadata of specified file _ read _ pointer defines new currency

Metadata _ write: the metadata that after the reception of metadata _ write command, is streamed to device is overwritten in the metadata of the offset address place specified file that the currency by metadata _ write _ pointer defines.The content of the metadata of specified file and length are confirmed by main frame.Metadata _ write command is ordered and is stopped through receiving any other.

Metadata _ read: the metadata at the specified file at the offset address place that the currency by metadata _ read _ pointer defines can be from installing fluidisation after receiving metadata _ reading order.When reaching the end of metadata, the metadata fluidisation stops, and this condition can be discerned by means of status command by main frame.Metadata _ reading order is ordered and is stopped through receiving any other.

Metadata _ write _ pointer: metadata _ write _ pointer commands is set to the offset address of appointment with the metadata of specified file _ write _ pointer.Metadata _ write _ pointer is followed metadata _ write command in metadata and is increased progressively by device when being streamed to device.

Metadata _ read _ pointer: metadata _ read _ pointer commands is set to the offset address of appointment with the metadata of specified file _ read _ pointer.Metadata _ read _ pointer is increased progressively by device when metadata is followed metadata _ reading order from the device fluidisation.

In some cases, main frame can have the hierarchical arrangement of object.For instance, file can be stored in catalogue and the sub-directory.Immediate data file rear end is stored the file that has no hierarchy (that is, being in smooth in logic layout) usually.In order to be in harmonious proportion two systems, the object protocol adaptor can be through using the hierarchy that produces institute's storage object with the metadata of object associated again.Keep at main frame under the situation of classification, about the information of the state of classification file when the file storage as metadata store.When the main frame access memory system, can at first read metadata, make the object protocol adaptor can confirm that hierarchy also can return to main frame with this information.In this way, even file is stored in the accumulator system under the situation of not considering this rating information, the object protocol adaptor also can produce rating information again.For instance, file be arranged in the catalogue and the sub-directory of the classification of main frame can be as metadata store when file storage.When main frame is attempted the content of access memory system, in turning back to the information of main frame, reflect catalogue and sub-directory information after a while.

Multi-protocol interface

In an embodiment of the present invention; Accumulator system is provided; It can receive and storage has the object of predetermined size (for example according to MTP), the file that receives through streamed files of the conduct of predetermined size, and the data sector with the logical address that on the logical address space that defines for accumulator system, defines.Come the configuration protocol adapter corresponding to these three kinds of agreements, and select protocol adaptor according to the agreement that main frame uses.

Figure 16 shows to have the accumulator system 629 of three protocol adaptors 632,632,634 that are connected to shared immediate data file interface 636.Immediate data file interface 636 and immediate data file storage device 638 are in the rear end 640 based on file, and said rear end 640 can be considered immediate data file rear end.Therefore, accumulator system 629 is used for from the shared rear end of the data of any one reception of said file interface.The subregion of storer maybe be dispensable in this system, thereby use the free space in the memory array effectively.Therefore, the block in the array can be used for being stored in the data that different time receives from file interface, object interface and LBA interface.Yet the data that receive from the LBA interface can be stored in this time only stores the block of the data that receive from the LBA interface.Manage from the data of file interface and object interface reception with the mode of common its file structure of reflection, thereby reduce refuse collection.

Figure 16 displaying is connected to the LBA interface 639 based on the rear end 640 of file through LBA protocol adaptor 634.LBA protocol adaptor 634 can be a logical file with the LBA data-switching, or to use certain alternate manner be to be used for the suitable form that receives by based on the rear end of file with the LBA data-switching.

Object interface 642 is connected to the rear end 640 based on file through object protocol adaptor 632.Substantially, object protocol adaptor 632 will arrive another agreement from the data and the command conversion of an agreement, to allow to use the rear end of the main frame access of object agreement based on file.Although this can only relate to from the object protocol command to based on the translating one to one of the order of file, in some cases, order does not have equivalent in another agreement.Under said situation, the object protocol adaptor possibly made than simply translating more thing.For instance, send the metadata that comprises object size at the MTP main frame and also then send under the situation of object, the end of object protocol adaptor identifying object also produces the response to the MTP main frame.The object protocol adaptor also produces the close file order to the rear end based on file at this moment.

File interface 644 is connected to the rear end 640 based on file through file protocol adapter 630.In some cases, use suitable agreement based on file main frame can with the rear end direct communication based on file, and must any not translating.Yet in other cases, main frame can use based on file but send file with agreement different protocol based on the rear end of file.In the case, file protocol adapter 630 is carried out translating of any necessity.

The accumulator system of Figure 16 and the compatible mainframe that uses at least three kinds of different agreements.If extra protocol adaptor is provided, so accumulator system also can with other compatible mainframe.

Figure 17 shows the accumulator system 629 of Figure 16, and wherein file protocol adapter 630, object protocol adaptor 632 and LBA protocol adaptor 634 are regarded as translation layer 750 jointly.Each protocol adaptor is provided at the particular host agreement and based on required translating between the rear end 640 of file.In some cases, accumulator system 629 can be at every turn and an above main-machine communication.For instance, accumulator system 629 can be connected to the network that is attached with a plurality of main frames.In other instance, individual host can have that to show as them seem the different application of the operation of different main frames, and can use different protocol to communicate by letter with accumulator system.In these cases, translation layer 750 must solve any conflict between the different main frames.Translation layer 750 provides the arbitration between the different main frames, and feasible rear end 640 based on file does not receive the conflict order.This possibly mean main frame access of refusal, till another main frame is accomplished particular task.For instance, if object protocol adaptor 632 just in transmission object or metadata, translation layer 750 can prevent that LBA protocol adaptor 634 from communicating by letter with the rear end 640 based on file with file protocol adapter 630 so, till operation is accomplished.This possibly mean to the main frame transmission busy signal of attempting via LBA interface 639 or file interface 634 access memory system.In some instances, different main frames possibly come the rear end 640 of access based on file through staggered affairs, makes it have the access in the identical time cycle.In these cases, translation layer 750 is arbitrated between main frame.

Figure 17 shows the interface layer 752 that contains file interface 644, object interface 642 and LBA interface 639.File interface 644, object interface 642 and LBA interface 639 are shown as distinct elements, yet this is logical expressions of showing the logic interfacing of accumulator system 629, and possibly not necessarily have three kinds of independent physical interfaces.In some instances, single physical interface (for example USB connector or SD connector) is shared for any other interface of file interface 644, object interface 642 and LBA interface 639 and use.Use which kind of interface and therefore use which kind of protocol adaptor to depend on the agreement that main frame uses.In some cases, when accumulator system was connected to main frame for the first time, main frame can be indicated host protocol as the part of the routine of shaking hands.In other cases, accumulator system 629 is inferred host protocol from the order that main frame sends.Interface layer 752 can detect employed host protocol from the indication of main frame transmission or with certain alternate manner, and selects the appropriate communication interface according to host protocol.Therefore, when accumulator system detects main frame and is just using MTP, alternative interface 642 and object protocol adaptor 632.Substantially, only select a protocol adaptor in any one time.Yet in some cases, but the alternate selection protocol adaptor to allow the interleaving access of two or more main frames.

Should be appreciated that the element of Figure 17 is corresponding to the logic module of accumulator system 629, and not necessarily corresponding to independent physical component.Therefore can realize maybe can using the suitable firmware of on controller, operating to realize the function of interface layer 752 and translation layer 750 through special circuit.In an example, the single physical interface is managed by Memory Controller, and Memory Controller is chosen in the protocol adaptor that moves on the Memory Controller so that data and command conversion are the form compatible with shared rear end.Although Figure 17 only shows three protocol adaptors, real accumulator system possibly have greater or less than three protocol adaptors.Possibly there is an above object protocol adaptor in some cases.For instance, possibly there is the object protocol adaptor that is used for MTP and be used for another object agreement (for example, PTP) independent object agreement.Similarly, possibly there is an above file protocol adapter that is used for different file protocols, and the Different L BA protocol adaptor that is used for Different L BA agreement.For instance, the title of Alan W.Sinclair is the instance that the 11st, 302, No. 764 patent application cases of " Logically-Addressed FileStorage Methods " provide substituting LBA agreement.

Conclusion

Although described various aspects of the present invention with respect to example embodiment of the present invention, should be appreciated that the present invention receives the interior protection of full breadth of appended claims.

Claims (27)

1. non-volatile memory device that connects removedly with different host apparatus, wherein said different host apparatus use different protocol in nonvolatile memory array, to store data, and said storage arrangement comprises:

Nonvolatile memory array; And

Controller, it is configured to:

First protocol adaptor; It receives first data from first application program as first Hosts file; Before said file was the indication of the said first Hosts file length; And said first data are sent to said nonvolatile memory array, wherein said first data storage is in the position of using first file identifier record;

Second protocol adaptor; It receives the indication that does not have the said second Hosts file length from second data of second application program as the data stream of the data that are identified as second Hosts file; And said second data are sent to said nonvolatile memory array, wherein said second data storage is in the position of using second file identifier record; And

The 3rd protocol adaptor; It receives the 3rd data from the 3rd application program as a plurality of sectors with the indivedual logical addresses that define on the logical address space that defines for said storage arrangement; And said the 3rd data are sent to said nonvolatile memory array, wherein said the 3rd data storage is in the position of using the 3rd file identifier record;

Wherein said first, second with the position of the 3rd data be with in the said nonvolatile memory array of indication corresponding to said first, second with the 3rd file identifier in each the clauses and subclauses of one or more blocks shine upon, to such an extent as to use identical agreement based on file will be said first, second and the 3rd data storage said nonvolatile memory array in.

2. non-volatile memory device according to claim 1, wherein said non-volatile memory device are included in to connect through standardization and are connected in the memory cards of host computer system with removable mode.

3. non-volatile memory device according to claim 2, wherein said first application program is moved on first host computer system, and said memory cards is connected to said first host computer system in the very first time:

Said second application program is moved on second host computer system, and said memory cards is connected to said second host computer system in second time; And

Said the 3rd application program is moved on the 3rd host computer system, and said memory cards is connected to said the 3rd host computer system in the 3rd time.

4. non-volatile memory device according to claim 1, wherein said first protocol adaptor produces the received indication of said first Hosts file to said main frame.

5. non-volatile memory device according to claim 1, wherein said first protocol adaptor produce the designator that said first Hosts file finishes, and said designator causes said first data of scheduling to carry out refuse collection.

6. non-volatile memory device that is presented as removable memory cards, said removable memory cards connects removedly with the different host apparatus that use different agreement and store the data through said HPI reception, and said storage arrangement comprises:

Nonvolatile memory array;

The back-end memory management system, its use is managed the data in the said nonvolatile memory array based on the agreement of file;

Interface layer, itself and main-machine communication; And

Translation layer; It is between said interface layer and said back-end memory management system, and said translation layer receives the Host Command from said interface layer, and said Host Command is observed the object agreement; And in response to receiving said Host Command; Said translation layer produces the order through translating to said back-end memory management system, does not saidly observe said object agreement through the order of translating

After wherein before main frame is sending object, sending the metadata of indication of the size that comprises said object, said translation layer is configured to confirm to receive whole object from said main frame according to said indication,

Said translation layer is used for receiving the Host Command of the said agreement based on file of using said back-end memory management system by further configuration, and as response, sends said Host Command to said back-end memory management system.

7. non-volatile memory device according to claim 6, wherein said object agreement are media transmission protocol (MTP).

8. non-volatile memory device according to claim 6; Wherein in response to confirming to have received said whole object; Said translation layer produces the response that sends to said main frame, and produces the end of file indicator that sends to said back-end memory management system.

9. non-volatile memory device according to claim 6, wherein said translation layer further comprises:

The file protocol adapter, it will be translated to said rear end file protocol from the communication of second main frame that uses the Hosts file agreement; And

The LBA protocol adaptor, it will be translated to said rear end file protocol from the communication of the 3rd main frame that uses the logical address agreement.

10. non-volatile memory device according to claim 9; Wherein said LBA protocol adaptor receives from the logical address space that defines for said accumulator system has the data sector by the logical address of said the 3rd host assignment, and with the virtual file of said sector map to the capacity of the first block that equals said memory array in size.

11. a non-volatile memory device that connects removedly with different host apparatus, wherein said different host apparatus use different protocol in nonvolatile memory array, to store data, and said storage arrangement comprises:

Nonvolatile memory array; And

Controller, it is configured to:

First protocol adaptor; It receives first data from first application program as first Hosts file; Before said file was the indication of the said first Hosts file length; And said first data are sent to said nonvolatile memory array, wherein said first data storage is in the position of using first file identifier record; And

Second protocol adaptor; It receives the indication that does not have the said second Hosts file length from second data of second application program as the data stream of the data that are identified as second Hosts file; And said second data are sent to said nonvolatile memory array; Wherein said second data storage is in the position of using second file identifier record

The position of wherein said first and second data uses identical agreement based on file, shines upon corresponding to the clauses and subclauses of the one or more blocks of each in said first and second file identifier in the said nonvolatile memory array through indicating.

12. non-volatile memory device according to claim 11, wherein said non-volatile memory device are included in to connect through standardization and are connected in the memory cards of host computer system with removable mode.

13. non-volatile memory device according to claim 12, wherein said first application program is moved on first host computer system, and said memory cards is connected to said first host computer system in the very first time: and

Said second application program is moved on second host computer system, and said memory cards is connected to said second host computer system in second time.

14. non-volatile memory device according to claim 11, wherein said first protocol adaptor produces the received indication of said first Hosts file to said main frame.

15. non-volatile memory device according to claim 11, wherein said first protocol adaptor produce the designator that said first Hosts file finishes, said designator causes said first data of scheduling to carry out refuse collection.

16. an operation comprises the method for the non-volatile memory device of block erasable memory array and controller; Said non-volatile memory device and a plurality of host protocol compatibility are being used for data storage in said block erasable memory array, and said method comprises:

Detect the host protocol when being connected to main frame, and from a plurality of protocol adaptors, select a protocol adaptor in response to detected said host protocol, said a plurality of protocol adaptors comprise first, second and the 3rd protocol adaptor at least;

Select said first protocol adaptor of said non-volatile memory device in response to detecting first host protocol; Said first host protocol sends the file of predetermined length; Before said file was the file size designator, and respective files has unique file indicator;

Select said second protocol adaptor of said non-volatile memory device in response to detecting second host protocol, said second host protocol sends the file that does not have the file size designator, and respective files has unique file indicator; And

Select said the 3rd protocol adaptor of said non-volatile memory device in response to detecting the 3rd host protocol; Said the 3rd host protocol sends data sector; Each sector has the logical address that defines on the logical address space that defines for said storage arrangement

First file storage with first file indicator that will receive through said first protocol adaptor and uses the agreement based on file to write down first relevance between said first file identifier and said more than first block in more than first block of said memory array;

Second file storage with second file indicator that will receive through said second protocol adaptor and uses said agreement based on file to write down second relevance between said second file identifier and said more than second block in more than second block of said memory array; And

A plurality of sector storage that will receive through said the 3rd protocol adaptor and use said agreement based on file to write down the 3rd relevance between the 3rd file identifier and said more than the 3rd block in more than the 3rd block.

17. method according to claim 16, wherein said first protocol adaptor produce the response to said first main frame after the data that receive said predetermined length.

18. method according to claim 16, wherein said first protocol adaptor produces end of file indicator in response to the data that receive said predetermined length, and said designator causes said file to be closed and warp is dispatched to carry out refuse collection.

19. method according to claim 16, it further is included in any one time and only selects one in said first, second or the 3rd protocol adaptor.

20. method according to claim 16, its further be included in different time alternately select said first, second with the 3rd protocol adaptor in not same person to allow the interleaving access of an above host application program.

21. method according to claim 16, wherein said accumulator system is implemented in the removable memory cards.

22. method according to claim 21, wherein said memory cards meets the standard that is selected from the following: the flash memory that compacts, multimedia card, secure digital, small-sized SD, memory stick, smart media or TransFlash.

23. an operation comprises the method for the removable non-volatile memory device of block erasable memory array and controller; Said non-volatile memory device and a plurality of host protocol compatibility are being used for data storage at said block erasable memory array, and said method comprises:

Detect the host protocol when being connected to main frame, and from a plurality of protocol adaptors, select a protocol adaptor in response to detected said host protocol, said a plurality of protocol adaptors comprise at least the first protocol adaptor and second protocol adaptor;

Select said first protocol adaptor of said non-volatile memory device in response to detecting first host protocol; Said first host protocol sends the file of predetermined length; Before said file was the file size designator, and respective files has unique file identifier; And

Select said second protocol adaptor of said non-volatile memory device in response to detecting second host protocol, said second host protocol sends the file that does not have the file size designator, and respective files has unique file identifier,

First file storage that will receive through said first protocol adaptor and uses the agreement based on file to write down the unique identifier of first file and first relevance between said more than first block in more than first block of said memory array; And

Second file storage that will receive through said second protocol adaptor and uses said agreement based on file to write down the unique identifier of second file and second relevance between said more than second block in more than second block of said memory array.

24. method according to claim 23, wherein said first protocol adaptor produce the response to said first main frame after the data that receive said predetermined length.

25. method according to claim 23, wherein said first protocol adaptor produces end of file indicator in response to the data that receive said predetermined length, and said designator causes said file to be closed and is scheduled to carry out refuse collection.

26. method according to claim 23, it further is included in any one time and only selects one in said first or second protocol adaptor.

27. method according to claim 23, it further comprises selects alternately one in first or second protocol adaptor, wherein said removable accumulator system and an above main-machine communication.

Images